A Diachronic Coherence Model for Language Program Evaluation



Language programs often generate complex date providing a rich source of comparative information about changes in curriculum policies and learning. Historical program archives often reveal how curricula and assessment methods evolve to include increasingly varied components of achievment. This article presents a novel quantitative methodology for assessing the construct of “program coherence” in an English as a Foreign Language context. Revolving panel studies demonstrate criteria for evaluating program coherence and its influence on proficiency gains. Thirty–three panel cohorts are included in a series of six 1–year longitudinal studies relating program–internal assement of achievement to program–external measures of language proficiency. Differences in panel coherence are analyzed in a hierarchical model that gauges the influence of coherence independently individual differences at the learner level.